Last updated: November 2026. This Privacy Policy explains how Prompt Reverse collects, uses, and protects your personal information.
We collect the minimum needed to run the platform: account information (name and email), uploaded content for processing (including songs, voice recordings, sound clips and videos with sound), the scripts, lyrics and voice descriptions you type into our audio tools, any voices you design, general usage data for security and service improvement, and payment information handled entirely by Stripe (we never see or store card details).
Files (images, videos, audio) are uploaded, sent to an AI model for processing, and the results are returned to you. Direct uploads are stored temporarily for processing and expire automatically after 15 minutes. Files you generate are kept in Recent generations for 7 days and then deleted automatically. Prompts and items you save to your library stay there until you delete them. We do not keep copies of your uploads unless you explicitly save results. Your uploads and outputs are never used to train, fine-tune, or improve any AI model.
When you reverse a voice, the recording is analysed to describe the voice in words, such as its tone, pace and accent. We do not create voiceprints or any other biometric identifier, we do not clone voices, and we do not try to identify who is speaking; the recording is only kept if you save it. Audio to Prompt describes the style of a song or sound and does not name artists, songs or speakers or transcribe sung lyrics. When you add subtitles, the speech in your video is converted to text. Scripts, lyrics and voice descriptions are checked against our content rules before anything is generated, and when we refuse a request we log the reason, your account and the tool so that abuse can be reviewed. A voice you design from a written description is a synthetic voice stored with our voice provider; you can keep up to three and delete any of them at any time in the AI Voice Generator, which removes the voice from your account and from the provider. Every audio file we generate is tagged as AI-generated in its metadata, and music made with Google's Lyria model also carries Google's SynthID watermark; these markings do not contain your name or email address.
We send content to AI providers including Anthropic (Claude), OpenAI (GPT, DALL-E), and Google (Gemini and Lyria), along with specialized image, video and audio providers such as xAI, Kling, Luma, ElevenLabs and Stability AI, via their APIs solely to generate your results. Some of these models, including Kling's video-to-audio model and Stability AI's Stable Audio, run through fal.ai, a platform that hosts AI models. For audio: ElevenLabs generates voiceovers, sound effects, designed voices and the music in launch videos, and converts speech to text for subtitles; Google Gemini listens to uploaded audio and video to analyse them and checks audio requests against our content rules; Google Lyria and Stable Audio generate music; and Kling generates sound to match a video. We use API access (not consumer products), so your data is not used to train their models. We do not share your account details with these providers.
Essential cookies for authentication and security, and analytics cookies used by Google Analytics to understand platform usage. You can opt out via browser settings.
You have full control: access what data we hold, correct your information, request deletion, or export your data at any time through your account settings. You can delete any voice you designed, at any time, in the AI Voice Generator.